US 7187036 B2
A semiconductor contact connection structure and the method for forming the same are disclosed. The connection structure has a first semiconductor device formed on an insulator substrate. A non-conducting gate interconnect layer is formed on the insulator substrate for connecting to a gate of a second semiconductor device, and a silicide layer formed on the gate interconnect layer and an active region of the first semiconductor device for making a connection thereof.
1. An SRAM cell formed on an insulator substrate, the cell comprising:
at least one active region formed on the insulator substrate and with a continuous silicide layer formed thereon serving as an intra-cell connection layer connecting at least a drain node of a PMOS transistor to a drain node of an NMOS transistor formed on the insulator substrate, the two transistors forming a first inverter;
said continuous silicide layer further forming a sidewall butted connection structure used in conjunction with a gate interconnect layer and connecting the drain nodes of the transistors of the first inverter to gates of at least two transistors of a second inverter,
wherein the continuous silicide layer consists of a first silicide film formed of silicon from the active region and a second silicide film formed of silicon from the gate interconnect layer.
2. The cell of
3. The cell of
4. The cell of
5. The cell of
6. The cell of
7. The cell of
8. An SRAM cell formed on an insulator substrate, the cell comprising:
a first inverter having a first PMOS transistor and a first NMOS transistor;
a second inverter having a second PMOS transistor and a second NMOS transistor; and
a sidewall butted connection structure used in conjunction with a gate interconnect layer for connecting drain nodes of the transistors of the first inverter formed on an active region disposed on the insulator substrate, to gates of the two transistors of the second inverter,
wherein the sidewall butted connection structure is a continuous silicide layer including a junction covering a dielectric edge portion between the active region and the gates of the two transistors and consists of a first silicide film formed of silicon from the active region and a second silicide film formed of silicon from the gates.
9. The cell of
10. The cell of
11. The cell of
12. The cell of
13. The cell of
14. The cell of
15. The cell of
16. The cell of
17. The cell of
18. The cell of
The present invention relates generally to semiconductor devices; and more particularly, to the contact structures for semiconductor devices. Still more particularly, the present invention relates to a device and process that uses sidewall butted connection and silicide bridge connection for reducing structure count, complexity, and size.
One type of basic semiconductor device is the CMOS static random access memory (SRAM) which retains its memory state without refreshing as long as power is supplied to the cells. In a typical SRAM, the memory is stored as a voltage differential within a bistable cell constructed of two cross-coupled inverters. Data is written into or read from the cell through two pass gate transistors oppositely biased by a bit line, and a bit bar line, and controlled by a word line. Each inverter is composed of a P-channel metal-oxide-semiconductor (MOS) field-effect-transistor (PMOSFET) in series with an input/output (I/O) node, and an N-channel MOSFET (NMOSFET). The node of each inverter is connected to the gates of both transistors of the other inverter. An I/O transistor is connected from a bit line to the node of a first inverter. Another I/O transistor is connected from a bit line bar (always biased oppositely from the Bit line) to the node of a second inverter. In semiconductor memory designs, large memory count, stable data retention, and speed are valued. However, large memory count and complex structures add up to a large cell physical size, which results in higher fabrication expense and slower speed. As such, reducing structure count of contacts and their complexity is relevant.
Efficient integrated circuit chip layout of a standard six-transistor SRAM may minimize the area required. However, as the demand for more complex integrated circuits, smaller transistors and structures, and for faster and more reliable performance continues to grow, new approaches are needed. Even the contact structures occupy valuable area in memory cells. Therefore, reductions in contact count, complexity, and size are of paramount importance.
Silicon-on-insulator (SOI) is a structure of silicon structure in which a thin layer of insulating oxide is buried just below the top surface. A thin layer of single-crystal silicon above the oxide may then be used for the construction of semiconductor devices. These devices are typically isolated in individual islands of the aforesaid thin layer of single-crystal silicon. By constructing and separating circuit elements with no substrate connection, latch-ups may be avoided in CMOS devices. Also, junction capacitance may be reduced, thereby achieving higher speed.
For maximum density to be achieved in logic circuits and SRAM, the cells must be laid out in as small a size as possible. In nanometer technology, the SOI MOSFETs have been researched for high-speed, low-leakage and high-density products. In addition, SOI technology allows for a tighter design rule because of well isolation margins.
Desirable in the art of semiconductor devices are additional designs and cell layouts that utilize SOI technologies in tandem with other technologies to achieve reduction in contact count, complexity and size, thereby improving overall speed and performance.
In view of the foregoing, this invention provides a circuit that reduces contact count, complexity and size.
In one example, a semiconductor contact connection structure and the method for forming the same are disclosed. The connection structure has a first semiconductor device formed on an insulator substrate. A non-conducting gate interconnect layer is formed on the insulator substrate for connecting to a gate of a second semiconductor device; and a silicide layer formed on the gate interconnect layer, and an active region of the first semiconductor device for making a connection thereof.
This improved contact structure efficiently and simply replaces a tungsten plug butted contact or a pair of contacts with a metal routing. This also eases the alignment tolerance for the contact. Various aspects and advantages will become apparent from the following detailed description, taken in conjunction with the accompanying drawings, illustrating the principles of the invention by way of examples.
The present invention provides an improved method for the layout of a silicon-on-insulator (SOI) version of an integrated circuit (IC), and a contact/connection structure therein. Several examples will be illustrated using a Static Random Access Memory (SRAM) cell, but it is understood that the same technology can be used for other semiconductor devices.
As will be illustrated below, the present invention uses a polysilicon (poly) sidewall butted connection structure, between the gate poly and active area, to replace a butted contact or a tungsten plug butted contact, or two contacts with a metal routing layer in the semiconductor device.
A metal (not shown), such as titanium, may be deposited on the entire top exposed surface of all illustrated structures. A high temperature process alloys the metal with all exposed silicon surfaces to produce a metal-silicide surface layer. Therefore, a metal-silicide 222 is formed on the surface of the gate poly layer 206. A metal-silicide 224 is formed on the surface of the source and drain regions. A metal-silicide 226 is formed on the surface of the connecting poly layer 210. A metal-silicide 228 is formed on the surface of the sidewall of the connecting poly layer 210. Excess unalloyed metal is then etched off. Since a metal that is selected will not alloy with silicon oxide, the remaining sidewall spacer 216 is cleared of metal. A dielectric edge portion such as a gate oxide layer edge 230 that is exposed when the sidewall spacer is etched from the sidewall of the connecting poly layer 210 may not be cleared of metal because the gap across the thickness of the gate oxide is relatively small (e.g., less than 50 angstroms). The metal-silicide may include TiSi2, CoSi2, NiSi, PtSi, WSi2, or other metal based materials such as TiN, TiW, TaN. Further, the metal-silicide surface layer may be less than 350 angstroms in thickness. In some applications, the metal-silicide surface layer provides an electrical resistance of 100 ohm/ea, or less, between the gate interconnect layer and the active region.
During the high temperature alloy process, not only does the metal diffuse into the silicon layers; but also, the silicon diffuses into the deposited metal layer, from both the single crystal silicon island 202 and the connecting poly layer 210. Metal-silicide is therefore formed in a continuous layer 232, even across the narrow gate oxide layer edge 230. Continuous layer 232 thereby bridges the gap formed by the edge of gate oxide 212, between single crystal silicon island 202 and connecting poly layer 210. Continuous layer 232 is formed by respective lateral encroachment of a silicide film formed of the silicidation of the metal layer and single crystal silicon island 202 and a silicide film formed of the silicidation of metal layer 202 and connecting poly layer 210. Continuous layer 232 thereby consists of a first silicide film that includes silicon from single crystal silicon island 202 and a second silicide film that includes silicon from connecting poly layer 210. As such, the gate oxide 212 no longer separates the silicon island 202 from the connecting poly silicon layer 210. This continuous silicide layer 232 is said to have a sidewall butted connection structure. From the perspective of an SRAM cell, this metal-silicide layer may serve as a local intra-cell connection mechanism between the pull-up P-channel metal-oxide-semiconductor field effect transistor (PMOSFET) and the pull-down NMOSFET, wherein the SOI P-active area is directly and continuously connected with the SOI N-active area.
Silicided P+ active areas include the source 320 and the drain 322 of the PMOSFET 314. Silicided N+ active areas include the source 324 and the drain 326 of the NMOSFET 316. A mask line 328 is the edge of the N-well that is implanted and diffused under the whole PMOSFET 314. N-well doping is overwhelmed everywhere except under the gate of PMOSFET 314 and, optionally, the portion of a silicided neck 330 between the P+ doped drain 322 and the mask line 328. The silicided neck 330 is the active area that may or may not be doped P-type or N-type. Doping is optional in the silicided neck 330, because the silicided surface shorts the silicided PMOSFET P+ doped drain 322 to the silicided NMOSFET N+ doped drain 326. Furthermore, the NMOSFET silicided N+ drain 326 is common with the silicided N+ source 332 of the pass gate NMOSFET 334 (PG-1), which has silicided N+ drain 336.
The connection, by the silicided neck 330, between the drains 322 and 326 obviates two contacts, one to each drain, that are necessary in previous technologies. The two drains 322 and 326, connected by the silicided neck 330, form the storage node of the inverter 318. The storage node is connected to the silicided gate poly 302 by the sidewall butted contact formed on the edge 306 of the silicided gate poly 302. This is the sidewall butted contact 232 as illustrated in
It is also understood that the inverter gate poly patterns can be extended so that they overlap the drains of both the NMOSFET and the PMOSFET. As such, the sidewall butted contacts can be formed for directly connecting the drains of the transistors of a first inverter to the gates of the transistors of a second inverter in the SRAM cell.
The above invention provides many different embodiments, or examples, for implementing different features of the invention. Specific examples of components, and processes are described to help clarify the invention. These are, of course, merely examples and are not intended to limit the invention from that described in the claims.
Although illustrative embodiments of the invention have been shown and described, other modifications, changes, and substitutions are intended in the foregoing invention. Accordingly, it is appropriate that the appended claims be construed broadly, and in a manner consistent with the scope of the invention, as set forth in the following claims.